Stock return prediction via multi-scale kernel adaptive filtering

doi:10.11772/j.issn.1001-9081.2022030401

Journal of Computer Applications ›› 2023, Vol. 43 ›› Issue (5): 1385-1393.DOI: 10.11772/j.issn.1001-9081.2022030401

Special Issue: 第九届中国数据挖掘会议(CCDM 2022)

• China Conference on Data Mining 2022 (CCDM 2022) • Previous Articles Next Articles

Stock return prediction via multi-scale kernel adaptive filtering

Xingheng TANG¹^,², Qiang GUO¹^,²(), Tianhui XU¹^,², Caiming ZHANG²^,³^,⁴

^1.School of Computer Science and Technology，Shandong University of Finance and Economics，Jinan Shandong 250014，China
^2.Shandong Provincial Key Laboratory of Digital Media Technology （Shandong University of Finance and Economics），Jinan Shandong 250014，China
^3.School of Software，Shandong University，Jinan Shandong 250101，China
^4.Shandong Provincial Laboratory of Future Intelligence and Financial Engineering （Shandong Technology and Business University），Yantai Shandong 264005，China

Received:2022-03-30 Revised:2022-05-18 Accepted:2022-05-30 Online:2023-05-08 Published:2023-05-10
Contact: Qiang GUO
About author:TANG Xingheng， born in 1998， M. S. candidate. His research interests include data mining， time-series data prediction.
GUO Qiang， born in 1979， Ph. D.， professor. His research interests include data mining， time-series data analysis， computer vision.
XU Tianhui， born in 1998， M. S. candidate. Her research interests include anomaly detection in time-series data.
ZHANG Caiming， born in 1955， Ph. D.， professor. His research interests include computer graphics， computer vision， medical image processing， time-series data analysis.
Supported by:
National Natural Science Foundation of China(61873145);Science and Technology Innovation Supporting Program for Distinguished Young Scholars of Shandong Province Higher Education Institutions(2019KJN045)

基于多尺度核自适应滤波的股票收益预测

汤兴恒¹^,², 郭强¹^,²(), 徐天慧¹^,², 张彩明²^,³^,⁴

^1.山东财经大学计算机科学与技术学院, 济南 250014
^2.山东省数字媒体技术重点实验室(山东财经大学), 济南 250014
^3.山东大学软件学院, 济南 250101
^4.山东省未来智能金融工程实验室(山东工商学院), 山东烟台 264005

通讯作者: 郭强
作者简介:汤兴恒（1998—），男，山东济宁人，硕士研究生，主要研究方向：数据挖掘、时序数据预测
郭强（1979—），男，山东淄博人，教授，博士，主要研究方向：数据挖掘、时序数据分析、计算机视觉 guoqiang@sdufe.edu.cn
徐天慧（1998—），女，山东临沂人，硕士研究生，主要研究方向：时序数据异常检测
张彩明（1955—），男，山东乳山人，教授，博士，主要研究方向：计算机图形学、计算机视觉、医学影像处理、时序数据分析。
基金资助:
国家自然科学基金资助项目(61873145);山东省高等学校青创科技支持计划项目(2019KJN045)

Abstract

Abstract:

In stock market， investors can predict the future stock return by capturing the potential trading patterns in historical data. The key issue for predicting stock return is how to find out the trading patterns accurately. However， it is generally difficult to capture them due to the influence of uncertain factors such as corporate performance， financial policies， and national economic growth. To solve this problem， a Multi-Scale Kernel Adaptive Filtering （MSKAF） method was proposed to capture the multi-scale trading patterns from past market data. In this method， in order to describe the multi-scale features of stocks， Stationary Wavelet Transform （SWT） was employed to obtain data components with different scales. The different trading patterns hidden in stock price fluctuations were contained in these data components. Then， the Kernel Adaptive Filtering （KAF） was used to capture the trading patterns with different scales to predict the future stock return. Experimental results show that compared with those of the prediction model based on Two-Stage KAF （TSKAF）， the Mean Absolute Error （MAE） of the results generated by the proposed method is reduced by 10%， and the Sharpe Ratio （SR） of the results generated by the proposed method is increased by 8.79%， verifying that the proposed method achieves better stock return prediction performance.

Key words: stock return prediction, Kernel Adaptive Filtering (KAF), trading pattern, multivariate data dependence, sequence learning

摘要：

在股票市场中，投资者可通过捕捉历史数据中潜在的交易模式实现对股票未来收益的预测，股票收益预测问题的关键在于如何准确地捕捉交易模式，但受公司业绩、金融政策以及国家经济增长等不确定性因素的影响，交易模式往往难以捕捉。针对该问题，提出一种多尺度核自适应滤波（MSKAF）方法，从过去的市场数据中捕捉多尺度交易模式。为刻画股票的多尺度特征，该方法采用平稳小波变换（SWT）得到不同尺度的数据分量，不同尺度的数据分量蕴含着股票价格波动背后潜在的不同交易模式，然后采用核自适应滤波（KAF）方法捕捉不同尺度的交易模式，以预测股票未来收益。实验结果表明，相较于基于两阶段核自适应滤波（TSKAF）的预测模型，所提方法的预测结果的平均绝对误差（MAE）减小了10%，夏普比率增加了8.79%，可见所提方法实现了更好的股票收益预测性能。

关键词: 股票收益预测, 核自适应滤波, 交易模式, 多元数据依赖, 序列学习

CLC Number:

TP181

Xingheng TANG, Qiang GUO, Tianhui XU, Caiming ZHANG. Stock return prediction via multi-scale kernel adaptive filtering[J]. Journal of Computer Applications, 2023, 43(5): 1385-1393.

汤兴恒, 郭强, 徐天慧, 张彩明. 基于多尺度核自适应滤波的股票收益预测[J]. 《计算机应用》唯一官方网站, 2023, 43(5): 1385-1393.

Figures/Tables 12

Fig. 1 Schematic diagram of MSKAF model

Fig. 2 Multi-level stationary wavelet decomposition

Fig. 3 Schematic diagram of outlier detection process

Fig. 4 Sampling process of stock time series

Fig. 5 Multivariate time series data dependence model

Tab. 1 Stock information used in experimental process

市场	股票代码	股票名称
DE	ADS	Adidas AG
	ALV	Allianz SE
	DPW	Deutsche Post AG
	DTE	Deutsche Telekom AG
	HEI	Heidelberg Cement AG
	LIN	Linde AG
	MRK	Merck KGaA
	SAP	SAP AG
UK	ADM	Admiral Group PLC
	AHT	Ashtead Group PLC
	BA	BAE Systems PLC
	BP	BP PLC
	CCL	Carnival PLC
	IAG	International Consolidated Airlines Group
	SKY	SKY PLC
	VOD	Vodafone Group PLC
US	AAL	American Airlines Group Inc
	AAPL	Apple Inc
	AMZN	Amazon Inc
	C	Citigroup Inc
	GOOGL	Alphabet In-CL A
	MSFT	Microsoft Corp
	SPY	SPDR S&P 500 Etf
	T	AT&T

Tab. 2 Parameter setting of comparison methods

参数	LSTM	VAR	VECM	NICE-KLMS	QKLMS	TSKAF	本文方法
时间窗口大小m	10	10	10	10	10	10	10
时间窗口滑动步长step	1	1	1	1	1	1	1
步长 $η$	—	—	—	0.05	0.05	0.05	0.05
带宽 $σ$	—	—	—	0.5	0.5	0.5	0.5
量化值 $λ$	—	—	—	0.06	0.4	—	—
质心距离 $β$	—	—	—	$2 σ$	—	—	—
异常点阈值 $δ$	—	—	—	—	—	10	—
低频序列异常点阈值 $δ l$	—	—	—	—	—	—	15
高频序列异常点阈值 $δ h$	—	—	—	—	—	—	10
信息量阈值 $ε$	—	—	—	—	—	0.000 1	0.000 1
网络层数 $L$	1	—	—	—	—	—	—
每层神经元数量 $N$	20	—	—	—	—	—	—
最大滞后次数 $ς$	—	15	—	—	—	—	—
滞后差异数 $A$	—	—	3	—	—	—	—
协整秩 $I$	—	—	0	—	—	—	—
超参数 $ν$	—	—	—	—	—	—	-6.5
超参数 $γ$	—	—	—	—	—	—	0.9

Tab. 2 Parameter setting of comparison methods

参数	LSTM	VAR	VECM	NICE-KLMS	QKLMS	TSKAF	本文方法
时间窗口大小m	10	10	10	10	10	10	10
时间窗口滑动步长step	1	1	1	1	1	1	1
步长 $η$	—	—	—	0.05	0.05	0.05	0.05
带宽 $σ$	—	—	—	0.5	0.5	0.5	0.5
量化值 $λ$	—	—	—	0.06	0.4	—	—
质心距离 $β$	—	—	—	$2 σ$	—	—	—
异常点阈值 $δ$	—	—	—	—	—	10	—
低频序列异常点阈值 $δ l$	—	—	—	—	—	—	15
高频序列异常点阈值 $δ h$	—	—	—	—	—	—	10
信息量阈值 $ε$	—	—	—	—	—	0.000 1	0.000 1
网络层数 $L$	1	—	—	—	—	—	—
每层神经元数量 $N$	20	—	—	—	—	—	—
最大滞后次数 $ς$	—	15	—	—	—	—	—
滞后差异数 $A$	—	—	3	—	—	—	—
协整秩 $I$	—	—	0	—	—	—	—
超参数 $ν$	—	—	—	—	—	—	-6.5
超参数 $γ$	—	—	—	—	—	—	0.9

Tab. 3 MAE values of different models in test phase

股票代码	LSTM	VAR	VECM	NICE-KLMS	QKLMS	TSKAF	本文方法
ADS	0.011 0	0.011 5	0.013 6	0.010 9	0.010 9	0.013 7	0.014 0
ALV	0.010 0	0.010 1	0.012 2	0.010 1	0.010 0	0.012 2	0.013 2
DPW	0.061 0	0.060 1	0.071 1	0.060 8	0.060 9	0.064 5	0.064 1
DTE	0.006 3	0.006 5	0.007 3	0.006 3	0.006 3	0.014 2	0.019 0
HEI	0.009 7	0.010 2	0.011 5	0.009 7	0.009 8	0.016 1	0.020 2
LIN	0.007 6	0.007 7	0.009 3	0.007 5	0.007 5	0.010 0	0.011 0
MRK	0.007 1	0.007 3	0.008 7	0.007 1	0.007 1	0.011 2	0.010 3
SAP	0.007 2	0.007 2	0.008 6	0.007 8	0.007 8	0.013 0	0.011 5
ADM	0.008 4	0.008 7	0.010 5	0.008 3	0.008 3	0.015 4	0.012 4
AHT	0.015 1	0.016 2	0.019 6	0.015 3	0.015 2	0.021 5	0.016 8
BA	0.009 1	0.009 3	0.010 6	0.008 7	0.008 8	0.013 4	0.016 7
BP	0.007 6	0.007 6	0.009	0.007 7	0.007 7	0.019 4	0.010 4
CCL	0.008 1	0.008 5	0.010 6	0.008 2	0.008 2	0.015 6	0.011 2
IAG	0.021 1	0.021 4	0.024 6	0.021 3	0.021 2	0.032 2	0.025 9
SKY	0.032 2	0.032 5	0.039 8	0.032 3	0.032 2	0.039 0	0.034 4
VOD	0.007 1	0.007 4	0.008 8	0.008 3	0.008 1	0.016 6	0.011 3
AAL	0.013 9	0.015 5	0.018 1	0.014 9	0.014 8	0.022 8	0.016 8
AAPL	0.008 5	0.008 8	0.009 5	0.008 6	0.008 5	0.022 4	0.026 1
AMZN	0.009 4	0.010 1	0.011 5	0.010 2	0.010 0	0.023 2	0.027 1
C	0.008 6	0.009 7	0.012 3	0.009 4	0.009 4	0.041 7	0.019 6
GOOGL	0.007 9	0.008 2	0.009 1	0.008 2	0.008 1	0.009 8	0.011 1
MSFT	0.007 6	0.007 8	0.009 0	0.007 4	0.007 5	0.010 0	0.012 1
SPY	0.003 9	0.004 1	0.005 0	0.003 8	0.003 8	0.005 7	0.006 8
T	0.008 2	0.008 1	0.009 6	0.008 1	0.008 2	0.010 5	0.011 2
均值	0.012	0.013	0.015	0.013	0.013	0.020	0.018
标准差	0.012	0.011	0.014	0.012	0.012	0.013	0.012

Tab. 4 MSE values of different models in test phase

股票代码	LSTM	VAR	VECM	NICE-KLMS	QKLMS	TSKAF	本文方法
ADS	0.000 3	0.000 3	0.000 4	0.000 3	0.000 3	0.000 4	0.000 4
ALV	0.000 3	0.000 3	0.000 3	0.000 3	0.000 3	0.000 3	0.000 4
DPW	0.013 4	0.013 4	0.017 2	0.013 5	0.013 2	0.014 0	0.013 8
DTE	0.000 1	0.000 1	0.000 1	0.000 1	0.000 1	0.000 3	0.000 3
HEI	0.000 2	0.000 2	0.000 2	0.000 2	0.000 2	0.000 4	0.000 4
LIN	0.000 1	0.000 1	0.000 1	0.000 1	0.000 1	0.000 2	0.000 2
MRK	0.000 1	0.000 1	0.000 2	0.000 1	0.000 1	0.000 2	0.000 2
SAP	0.000 1	0.000 1	0.000 1	0.000 1	0.000 1	0.000 3	0.000 2
ADM	0.000 2	0.000 2	0.000 2	0.000 2	0.000 2	0.000 4	0.000 3
AHT	0.000 5	0.000 5	0.000 7	0.000 5	0.000 5	0.000 8	0.000 6
BA	0.000 2	0.000 2	0.000 2	0.000 2	0.000 2	0.000 3	0.000 4
BP	0.000 1	0.000 1	0.000 1	0.000 1	0.000 1	0.000 5	0.000 2
CCL	0.000 1	0.000 1	0.000 2	0.000 1	0.000 1	0.000 4	0.000 2
IAG	0.000 9	0.000 9	0.001 1	0.000 9	0.000 9	0.001 6	0.001 2
SKY	0.002 8	0.002 8	0.003 7	0.002 8	0.002 8	0.003 4	0.003 0
VOD	0.000 1	0.000 1	0.000 2	0.000 1	0.000 1	0.000 4	0.000 2
AAL	0.000 4	0.000 4	0.000 5	0.000 4	0.000 4	0.000 8	0.000 5
AAPL	0.000 2	0.000 2	0.000 2	0.000 2	0.000 2	0.000 6	0.000 6
AMZN	0.000 2	0.000 2	0.000 3	0.000 2	0.000 2	0.000 7	0.000 6
C	0.000 1	0.000 2	0.000 2	0.000 2	0.000 2	0.001 9	0.000 5
GOOGL	0.000 1	0.000 1	0.000 2	0.000 1	0.000 1	0.000 2	0.000 2
MSFT	0.000 1	0.000 1	0.000 2	0.000 1	0.000 1	0.000 2	0.000 2
SPY	0.000 0	0.000 0	0.000 1	0.000 0	0.000 0	0.000 1	0.000 1
T	0.000 1	0.000 1	0.000 2	0.000 1	0.000 1	0.000 2	0.000 2
均值	0.001	0.001	0.001	0.001	0.001	0.001	0.001
标准差	0.003	0.003	0.004	0.003	0.003	0.003	0.003

Tab. 5 SR values of different models in test phase

股票代码	LSTM	VAR	VECM	NICE-KLMS	QKLMS	TSKAF	本文方法
ADS	0.901 7	2.268 2	0.936 9	0.216 2	0.005 1	3.651 7	4.240 6
ALV	1.296 5	1.081 7	0.002 1	0.208 9	0.072 9	3.392 1	4.770 8
DPW	-0.752 2	-0.871 7	0.794 8	-1.002 7	-1.118 7	2.126 2	-0.183 0
DTE	0.268 8	0.766 5	1.257 1	-0.037 2	-0.123 6	1.946 3	5.290 4
HEI	0.857 4	1.390 8	2.007 7	0.208 9	-0.028 7	2.005 1	5.726 7
LIN	0.504 4	1.262 3	2.507 4	0.112 6	0.064 6	3.122 8	4.678 0
MRK	0.729 6	1.465 5	0.675 6	0.098 3	-0.025 7	3.620 5	4.125 8
SAP	0.740 5	1.860 4	2.480 6	0.049 7	-0.037 1	4.382 1	5.416 0
ADM	0.466 3	2.432 6	1.016 4	0.334 0	0.006 5	5.951 0	6.663 2
AHT	2.279 4	-1.324 6	-0.377 8	0.936 0	-1.204 8	5.361 4	2.932 4
BA	0.724 6	1.209 7	1.265 6	0.099 8	-0.057 9	2.195 2	5.177 1
BP	1.036 1	-0.027 5	-1.303 4	1.564 5	0.109 4	6.537 3	5.540 4
CCL	1.243 3	-0.743 5	-0.515 4	0.545 6	0.021 6	5.976 4	6.126 7
IAG	1.919 3	0.366 2	-1.349 7	-1.123 0	1.566 2	4.557 0	-0.644 6
SKY	-0.246 9	-1.289 8	-1.333 4	0.363 7	0.318 0	1.424 8	0.749 3
VOD	0.695 5	0.389 1	1.422 7	0.222 0	0.168 8	5.249 1	6.593 1
AAL	1.131 9	-1.144 0	-1.161 6	-0.577 4	0.494 6	5.882 0	3.847 8
AAPL	0.643 2	1.042 2	1.648 0	0.127 9	-0.127 4	2.664 4	5.001 1
AMZN	0.593 4	1.032 7	1.730 7	0.168 1	-0.020 5	1.549 9	3.922 9
C	0.975 0	1.655 6	-0.303 1	0.273 7	0.087 9	3.655 5	-2.387 9
GOOGL	0.511 8	1.333 4	1.283 5	0.118 5	0.045 9	2.971 0	4.546 8
MSFT	0.529 1	1.226 8	2.110 7	0.130 1	0.035 5	2.888 1	4.535 2
SPY	0.412 1	1.071 1	1.725 9	0.082 7	-0.097 7	1.784 6	3.073 8
T	1.294 7	1.368 1	1.182 8	0.143 8	-0.068 4	3.339 6	4.072 6
均值	0.770	0.722	0.738	0.136	0.004	3.593	3.909
标准差	0.606	1.061	1.207	0.517	0.484	1.562	2.319

Tab. 6 Influence of multi-scale decomposition layers on experimental results

分解层数k	SR
1	2.019
2	3.909
3	0.305

Fig. 6 Influence of m and step on experimental results

References 34

1	OH K J， KIM K J. Analyzing stock market tick data using piecewise nonlinear model［J］. Expert Systems with Applications， 2002， 22（3）： 249-255. 10.1016/s0957-4174(01)00058-6
2	ARIYO A A， ADEWUMI A O， AYO C K. Stock price prediction using the ARIMA model［C］// Proceedings of the UKSim-AMSS 16th International Conference on Computer Modelling and Simulation. Piscataway： IEEE， 2014： 106-112. 10.1109/uksim.2014.67
3	TSAY R S. Analysis of Financial Time Series［M］. Hoboken， NJ： John Wiley & Sons， Inc.， 2005： 134-135. 10.1002/0471746193
4	FISCHER T， KRAUSS C. Deep learning with long short-term memory networks for financial market predictions［J］. European Journal of Operational Research， 2018， 270（2）： 654-669. 10.1016/j.ejor.2017.11.054
5	GARCIA-VEGA S， ZENG X J， KEANE J. Stock returns prediction using kernel adaptive filtering within a stock market interdependence approach［J］. Expert Systems with Applications， 2020， 160： No.113668. 10.1016/j.eswa.2020.113668
6	ADAM K， MARCET A， NICOLINI J P. Stock market volatility and learning［J］. The Journal of Finance， 2016， 71（1）： 33-82. 10.1111/jofi.12364
7	谭绮君. 基于多尺度分解和LSTM网络的金融时间序列预测研究［D］. 广州：华南理工大学， 2020： 1-8.
	TAN Q J. Financial time series prediction based on multiscale decomposition and long short-term memory networks［D］. Guangzhou： South China University of Technology， 2020： 1-8.
8	ZHANG L H， AGGARWAL C， QI G J. Stock price prediction via discovering multi-frequency trading patterns［C］// Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York： ACM， 2017： 2141-2149. 10.1145/3097983.3098117
9	REJEB A B， ARFAOUI M. Financial market interdependencies： a quantile regression analysis of volatility spillover［J］. Research in International Business and Finance， 2016， 36： 140-157. 10.1016/j.ribaf.2015.09.022
10	SIMS C A. Macroeconomics and reality［J］. Econometrica， 1980， 48（1）： 1-48. 10.2307/1912017
11	SIMS C A. Comparison of interwar and postwar business cycles： monetarism reconsidered［J］. The American Economic Review， 1980， 70（2）： 250-257.
12	ENGLE R F， GRANGER C W J. Co-integration and error correction： Representation， estimation， and testing［J］. Econometrica， 1987， 55（2）： 251-276. 10.2307/1913236
13	VALENTE J M， MALDONADO S. SVR-FFS： a novel forward feature selection approach for high-frequency time series forecasting using support vector regression［J］. Expert Systems with Applications， 2020， 160： No.113729. 10.1016/j.eswa.2020.113729
14	韩敏，马俊珠，任伟杰，等. 基于核自适应滤波器的时间序列在线预测研究综述［J］. 自动化学报， 2021， 47（4）： 730-746.
	HAN M， MA J Z， REN W J， et al. A survey of time series online prediction based on kernel adaptive filters［J］. Acta Automatica Sinica， 2021， 47（4）： 730-746.
15	LIU W F， POKHAREL P P， PRINCIPE J C. The kernel least-mean-square algorithm［J］. IEEE Transactions on Signal Processing， 2008， 56（2）： 543-554. 10.1109/tsp.2007.907881
16	CHEN B D， ZHAO S L， ZHU P P， et al. Quantized kernel least mean square algorithm［J］. IEEE Transactions on Neural Networks and Learning Systems， 2012， 23（1）： 22-32. 10.1109/tnnls.2011.2178446
17	LI K， PRINCIPE J C. Transfer learning in adaptive filters： the nearest instance centroid-estimation kernel least-mean-square algorithm［J］. IEEE Transactions on Signal Processing， 2017， 65（24）： 6520-6535. 10.1109/tsp.2017.2752695
18	SEZER O B， GUDELEK M U， OZBAYOGLU A M. Financial time series forecasting with deep learning： a systematic literature review： 2005-2019［J］. Applied Soft Computing， 2020， 90： No.106181. 10.1016/j.asoc.2020.106181
19	赵洪科，吴李康，李徵，等. 基于深度神经网络结构的互联网金融市场动态预测［J］. 计算机研究与发展， 2019， 56（8）： 1621-1631. 10.7544/issn1000-1239.2019.20190330
	ZHAO H K， WU L K， LI Z， et al. Predicting the dynamics in Internet finance based on deep neural network structure［J］. Journal of Computer Research and Development， 2019， 56（8）： 1621-1631. 10.7544/issn1000-1239.2019.20190330
20	杜圣东，李天瑞，杨燕，等. 一种基于序列到序列时空注意力学习的交通流预测模型［J］. 计算机研究与发展， 2020， 57（8）： 1715-1728. 10.7544/issn1000-1239.2020.20200169
	DU S D， LI T R， YANG Y， et al. A sequence-to-sequence spatial-temporal attention learning model for urban traffic flow prediction［J］. Journal of Computer Research and Development， 2020， 57（8）： 1715-1728. 10.7544/issn1000-1239.2020.20200169
21	MESHRAM S G， GHORBANI M A， SHAMSHIRBAND S， et al. River flow prediction using hybrid PSOGSA algorithm based on feed-forward neural network［J］. Soft Computing， 2019， 23（20）： 10429-10438. 10.1007/s00500-018-3598-7
22	RATHER A M， AGARWAL A， SASTRY V N. Recurrent neural network and a hybrid model for prediction of stock returns［J］. Expert Systems with Applications， 2015， 42（6）： 3234-3241. 10.1016/j.eswa.2014.12.003
23	HOCHREITER S， SCHMIDHUBER J. Long short-term memory［J］. Neural Computation， 1997， 9（8）： 1735-1780. 10.1162/neco.1997.9.8.1735
24	YAMANISHI K， TAKEUCHI J. A unifying framework for detecting outliers and change points from non-stationary time series data［C］// Proceedings of the 8th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York： ACM， 2002： 676-681. 10.1145/775047.775148
25	TAKEUCHI J， YAMANISHI K. A unifying framework for detecting outliers and change points from time series［J］. IEEE Transactions on Knowledge and Data Engineering， 2006， 18（4）： 482-492. 10.1109/tkde.2006.1599387
26	SCHÖLKOPF B， SMOLA A J. Learning with Kernels： Support Vector Machines， Regularization， Optimization， and Beyond［M］. Cambridge： MIT Press， 2002： 25-48.
27	SIIKANEN M， BALTAKYS K， KANNIAINEN J， et al. Facebook drives behavior of passive households in stock markets［J］. Finance Research Letters， 2018， 27： 208-213. 10.1016/j.frl.2018.03.020
28	LÜTKEPOHL H. Vector autoregressive models［M］// HASHIMZADE N， THORNTON M. Handbook of Research Methods and Applications in Empirical Macroeconomics. Northampton， MA： Edward Elgar Publishing， 2013： 139-164. 10.4337/9780857931023
29	KUO C Y. Does the vector error correction model perform better than others in forecasting stock price？ an application of residual income valuation theory［J］. Economic Modelling， 2016， 52（Pt B）： 772-789. 10.1016/j.econmod.2015.10.016
30	BAEK Y， KIM H Y. ModAugNet： a new forecasting framework for stock market index value with an overfitting prevention LSTM module and a prediction LSTM module［J］. Expert Systems with Applications， 2018， 113： 457-480. 10.1016/j.eswa.2018.07.019
31	KIRAN R， KUMAR P， BHASKER B. DNNRec： a novel deep learning based hybrid recommender system［J］. Expert Systems with Applications， 2020， 144： No.113054. 10.1016/j.eswa.2019.113054
32	WANG W Y， LI W Z， ZHANG N， et al. Portfolio formation with preselection using deep learning from long-term financial data［J］. Expert Systems with Applications， 2020， 143： No.113042. 10.1016/j.eswa.2019.113042
33	CUI Y W， AHMAD S， HAWKINS J. Continuous online sequence learning with an unsupervised neural network model［J］. Neural Computation， 2016， 28（11）： 2474-2504. 10.1162/neco_a_00893
34	LIU W F， PRINCIPE J C， HAYKIN S. Kernel Adaptive Filtering： A Comprehensive Introduction［M］. Hoboken， NJ： John Wiley & Sons， Inc.， 2010： 34-37. 10.1002/9780470608593

Stock return prediction via multi-scale kernel adaptive filtering

基于多尺度核自适应滤波的股票收益预测

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 12

References 34

Related Articles 1

Recommended Articles

Metrics